Vous êtes sur la page 1sur 32

COANDA HYDRO INTAKE SCREEN

TESTING AND EVALUATION

ETSU H/06/00053/REP

DTI/Pub URN 01/688

Contractor
Dulas Ltd

Prepared by
J Howarth

The work described in this


report was carried out under
contract as part of the New &
Renewable Energy
Programme, managed by
ETSU on behalf of the
Department of Trade and
Industry. The views and
judgements expressed in this
report are those of the
contractor and do not
necessarily reflect those of
ETSU or the Department of
Trade and Industry.

First published 2001


Crown copyright 2001

i
1 EXECUTIVE SUMMARY

1.1 Project Objectives


The objective of this project has been to evaluate the effectiveness, suitability
and cost benefit of the Aquashear Coanda effect, maintenance free intake
screen for use in small hydro system intakes.

1.2 Background and Project Need


High operational costs, reduced energy output and turbine wear are common,
but often overlooked aspects of conventional hydro intake and desilting
systems. Maintenance schedules, even for a well designed system, are often
not followed, resulting in costly remedial action and loss of generation. The
screen tested here offers an effective solution to these problems.

An existing small hydro system, near Keswick in Cumbria, UK has been used
for the project. A new Coanda screen was installed at the site and
commissioned in April 1999. Prior to the installation of the new screen, there
was an existing conventional intake that suffered from problems of blockage
with leaves and debris and was also unable to remove silt from the water that
passes through the turbine.

The Aquashear Coanda effect intake screen consists of parallel, horizontal


wedge wires, spaced at 1mm intervals. The screen excludes all debris greater
than 1mm diameter and 90% of particles down to 0.5mm, while achieving a
high flow capacity per meter width of screen. This is achieved by exploiting
two effects; the shearing action of the sharp wedge wires and the Coanda
effect, which is the tendency of fluids to cling to a surface. In this way, the
flow is drawn through the screen at a high rate.

The screen has been installed at many sites in the United States, primarily for
drinking water, irrigation and fish schemes. To date, Dulas1 has supplied
screens for installation at 15 locations in the UK and Europe for hydro
schemes. Until this study, there had been no extensive monitoring of the
performance of the screens after installation.

The results of the project will provide information to developers and operators
of small hydro plant for the design and refurbishment of intake structures.
The performance evaluation of the screen after 15 months in service has
resulted in a valuable cost-benefit analysis for the screen, compared to
conventional designs. Use of the screen should show significant increases in
energy output and hence revenue for operators. Dulas have identified markets
for the screen in Europe, Indonesia and India.

1
For more details, contact Dulas Ltd, Hydro, Dyfi Eco Parc, Machynlleth, Powys, SY20 8AX

ii
1.3 Work Carried Out
Data has been collected on the performance of the screen in this installation,
over a 15 month period, since the time of commissioning. The investigation
included analysis of the following performance parameters:

1.3.1 Screen Capacity.

The flow capacity of the screen was tested, both at the start and the end of the
15 month monitoring period, to determine if wear and abrasion in service had
reduced performance. The wedge wire edges of the screen were examined
with a microscope to determine if there was any visual signs of wear. To put
these results into perspective, samples of silt were taken from the river bed
and analysed to determine their abrasive properties.

1.3.2 Silt Exclusion Performance.

Samples of silt were washed over the screen and the amount of silt passing
through was assessed in terms of particle size and quantity.

1.3.3 Self Cleaning Operation.

Monitoring of river flows and turbine flows was carried out to determine if the
screen had become blocked at any time. The screen was also inspected at
regular intervals to look for signs of blockage.

1.3.4 Slime and Algae Growth.

The screen was regularly inspected to monitor signs of build up of algae. The
capacity of the screen was assessed to determine if the algae had any
deleterious effect. Samples of the water were taken and analysed for pH and
the nutrients nitrate and phosphate, to determine if there were favourable
conditions for algal growth.

1.3.5 Operation and Maintenance Requirements.

Any operating and maintenance requirements for the screen were recorded on
the daily site register.

1.3.6 Integrity and Resistance to Damage.

The screen was carefully inspected, particularly after high flood flows, to
check for signs of impact damage from boulders or tree trunks etc.

iii
1.3.7 Cost Benefit Analysis.

As a result of the above investigation, it was possible to carry out a cost


benefit analysis for the use of the screen at this site. This was achieved by
comparing capital costs, operation and maintenance costs, and energy output
and availability of the turbine and the resulting revenue. The use of the
Aquashear screen was compared with a conventional screen to give a Net
Present Value (NPV) for each type of screen. The NPV calculations were
carried out at 2.5% inflation and 8.5% discount rate.

1.4 Summary of Results

1.4.1 Screen Capacity.

The capacity of the screen was found to be approximately four percent less
than when installed. However, the errors present in accurately measuring the
flow were estimated at around 3%, hence it is difficult to draw a definite
conclusion. Microscopic examination of the screen wire edges showed no
visible signs of wear. The mineralogical analysis of typical silt found in the
river system, did not show any significant presence of highly abrasive
material. Hence it is thought unlikely that there would be a significant
deterioration of performance due to wear of the screen material.

There was however, evidence of the build up a limited amount of biological


growth, probably algae. This growth had coated part of the screen and it is
possible that this could reduce capacity slightly. The effects of this growth
can be mitigated by cleaning, and this is discussed more fully below.

1.4.2 Silt Exclusion Performance.

Several tests were carried out over a range of flows. At low flows, where
there was insufficient water to produce an excess leaving the bottom of the
screen, the silt exclusion performance was lower, averaging around 50%
exclusion of particles between 0.41 and 1.17mm. At higher flow tests, much
better performance was achieved, with average exclusion rates for the same
particle range of 94%. It was felt that this was a more indicative test, since
large volumes of small size silt particles would only be in suspension at high
flows. In such instances, there is sufficient excess flow to remove silt
particles from the screen, rather than allowing them to sit on the surface and
trickle through the wedge wires.

1.4.3 Self Cleaning Operation and Performance in Freezing Conditions.

iv
Neither the flow records for the turbine, or the operator reports showed any
evidence of blockage of the screen by debris. The operator reported that he
had never had to remove any debris or blockage from the screen.

It has been discovered however, that very thin strands of weed are able to pass
in small quantities through the wedge wires. This problem manifested itself in
a reduction in turbine maximum output power of about 7kW, or about four
percent. This occurred very gradually over a period of 15 months until the
operator decided to investigate the problem. Upon opening up the spear valve,
thin strands of weed were discovered. After cleaning these out, the turbine
output returned to normal. This was a short maintenance operation that only
required about an hour to carry out.

The winter of 1999-2000 was not particularly cold, so the screen was not
tested under severe freezing conditions. However, there was a few occasions
when the air temperature reached freezing point, and a period of several days
where it remained between zero degrees and minus two degrees celsius. No
problems of freezing of the screen or blockage by ice were noted.

1.4.4 Slime and Algae Growth

During the 15 month period, it was noticed that a thin film of algae had started
to spread over the screen. After about 12 months of operation, it was noted
that this had reduced the capacity of the screen. The screen was cleaned with
a stiff brush and the capacity of the screen returned to normal. Again this was
an isolated maintenance operation that only required about 1 hour to perform.

The water quality analysis showed that although the pH of the water was
conducive to algae growth, the concentrations of the nitrate and phosphate
were low, hence, higher than normal growth rates of algae would not be
expected.

1.4.5 Operation and Maintenance Requirements.

Apart from the two occasions described above, no maintenance was required
on the scheme due to problems with the screen.

1.4.6 Integrity and Resistance to Damage

The screen was inspected after several high floods where large tree trunks and
boulders had been washed over the weir. There was no evidence of any
damage to the screen. This is mainly due the orientation of the screen, where
objects tend to roll over it rather than impact with high force. This resistance
to damage has been observed at other installations, after more than 5 years in
service.

v
1.4.7 Cost Benefit Analysis

The NPV calculations carried out for the Coanda screen and a conventional
screen showed a significant advantage to using the Coanda screen. Over 5
years, there was a 14,000 benefit and over 10 years this amounted to
30,000. The majority of this benefit arises from the extra energy production
possible by using a self cleaning screen, that never blocks, compared to a
conventional screen can be 50% blocked for the whole of the autumn, when
high leaf loads are present.

1.5 Conclusions

From the results obtained from this investigation, it is clear that the
installation of a Coanda screen at this location has been highly successful,
resulting in payback on the installation costs within 1-2 years. The screen has
been effectively 100% reliable, since there has been no lost energy production
due to blockage by debris.

The screen has proven to be of consistent high capacity, robust, resistant to


damage from large boulders and tree trunks, unaffected by freezing
temperatures and has a high silt exclusion performance. The capacity tests
showed no significant deterioration in performance over a 15 month period.
The screens are also supplied approximately 50% oversized, hence even in
more aggressive, abrasive environments, the screens will be capable of
exceeding the performance requirements over a long design life.

Maintenance requirements have been negligible, with a total of about 3 hours


over 15 months being required for screen related maintenance activities. The
investigation has however, highlighted two minor areas where more careful
monitoring of performance may be required. The first is monitoring of algal
growth, which, if allowed to build up over long periods, can reduce the
capacity of the screen. In nutrient enriched waters, it may be necessary to
clean the screen once a month. The second area is monitoring of turbine
power output to check for build up of weed that has passed through the screen
and collected on the turbine spear valve. Both of these processes occur
relatively slowly, hence the cleaning can be a planned maintenance activity,
during a dry period. Hence, no loss of electricity production need occur.

Over the long term, the installation of an Aquashear Coanda screen is an


extremely good investment. For this particular site, the projected extra Net
Present Value from installation of the Coanda screen is 30,000 over a 10 year
period. For sites where a settling tank would be required in addition to the
conventional screen, the extra Net Present Value over 10 years was 38,000.
On some schemes, this extra availability and generated revenue, provided by
the Coanda screen, could be sufficient to turn an economically unattractive
project into a commercial viability.

vi
1.6 Recommendations

The screen has been tested over a period of 15 months for a range of
performance parameters. The screen has performed well in all of these tests.
The conclusions for most of the tests can be easily and accurately extrapolated
to the long term. For instance, tests for performance under freezing
conditions, resistance to blockage by debris, impact damage resistance and silt
exclusion should not be time dependent. However, there are two performance
parameters that would be of interest to investigate over the longer term;

1) Reduction of capacity due to build up algae or biological growth. It would


be very informative to monitor the build up of algae after a longer period
of 3 or 5 years. During this time it would be expected that the screen
would continually accumulate algae, which would be removed at intervals
by the operator as necessary. The current method is to use a stiff brush to
remove the growth from the outside screen surface. An increase in
capacity back to normal is usually noted. However, this brushing does not
return the screen to bare metal condition. It is possible that over longer
periods, more resistant residues may accumulate which could require a
more intensive cleaning process.

2) Reduction in capacity due to abrasion and blunting of the wedge wire


edges. The river in which the screen was installed did not contain large
volumes of abrasive material. A more meaningful test to determine long
term performance and life, would be to test the capacity of the screen after
5 years. An alternative would be to carry out a perfomance monitoring
programme on a screen installed in a more silty, abrasive environment. A
screen has recently been installed in Sri Lanka, which experiences high silt
loading. As the site is also in a tropical climate the rate of growth of algae
is much faster than for sites in the UK. This could therefore be a useful
site to monitor these two parameters of abrasion and algae growth.

The Aquashear Coanda screen is also supplied with smaller wedge wire
spacings. It would therefore be interesting to test a screen with 0.5mm
aperture. This type of screen would have many applications for higher head
sites, where the removal of smaller silt particles is essential to minimise
abrasion. It would be possible to do this on a small scale, by installing a
small, prefabricated section of screen in an appropriate river and monitoring
algae growth, abrasion and capacity over a 3 or 5 year period. The screen
would not necessarily have to be part of a hydro scheme to test these
parameters.

vii
1 EXECUTIVE SUMMARY ................................................................................... I

1.1 PROJECT OBJECTIVES .................................................................................... II


1.2 BACKGROUND AND PROJECT NEED .............................................................. II
1.3 WORK CARRIED OUT .................................................................................... III
1.3.1 SCREEN CAPACITY.....................................................................................III
1.3.2 SILT EXCLUSION PERFORMANCE...............................................................III
1.3.3 SELF CLEANING OPERATION. ....................................................................III
1.3.4 SLIME AND ALGAE GROWTH. ....................................................................III
1.3.5 OPERATION AND MAINTENANCE REQUIREMENTS. ...................................III
1.3.6 INTEGRITY AND RESISTANCE TO DAMAGE................................................III
1.3.7 COST BENEFIT ANALYSIS. ........................................................................ IV
1.4 SUMMARY OF RESULTS ................................................................................. IV
1.4.1 SCREEN CAPACITY.................................................................................... IV
1.4.2 SILT EXCLUSION PERFORMANCE.............................................................. IV
1.4.3 SELF CLEANING OPERATION AND PERFORMANCE IN FREEZING
CONDITIONS. ............................................................................................................... IV
1.4.4 SLIME AND ALGAE GROWTH ......................................................................V
1.4.5 OPERATION AND MAINTENANCE REQUIREMENTS. ....................................V
1.4.6 INTEGRITY AND RESISTANCE TO DAMAGE .................................................V
1.4.7 COST BENEFIT ANALYSIS ......................................................................... VI
1.5 CONCLUSIONS ................................................................................................ VI
1.6 RECOMMENDATIONS .................................................................................... VII

2 TECHNICAL BACKGROUND .......................................................................... D

2.1 DESCRIPTION OF AQUASHEAR SCREEN ........................................................ D


2.2 BENEFITS OF THE AQUASHEAR COANDA SCREEN ........................................ E
2.2.1 SILT REMOVAL .......................................................................................... E
2.2.2 ORGANIC MATERIAL REMOVAL ................................................................ E
2.2.3 EFFECT ON FLORA AND FAUNA ................................................................. E
2.2.4 POTENTIAL MARKETS .................................................................................F
2.3 PROJECT LOCATION ....................................................................................... F

3 METHODOLOGY OF ASSESSMENT AND RESULTS ................................ G

3.1 SCREEN CAPACITY AND WEAR ......................................................................G


3.1.1 METHOD..................................................................................................... G
3.1.2 RESULTS AFTER COMMISSIONING .............................................................. H
3.1.3 RESULTS AFTER 15 MONTHS OPERATION................................................... H
3.1.4 WEAR ...........................................................................................................I
3.1.5 MINERALOGICAL ANALYSIS OF SILT .......................................................... J
3.1.6 DISCUSSION................................................................................................. J
3.2 SILT REMOVAL ................................................................................................K
3.2.1 METHOD..................................................................................................... K
3.2.2 SILT EXCLUSION RESULTS ......................................................................... K
3.2.3 ANALYSIS OF SILT EXCLUSION PERFORMANCE FOR LOW FLOW TESTS ... L
3.2.4 HIGHER FLOW TESTS ................................................................................. L
3.3 SELF CLEANING OPERATION AND OPERATION IN FREEZING CONDITIONS
M
3.3.1 METHOD.................................................................................................... M

A
3.3.2 OPERATOR REPORTS .................................................................................. N
3.3.3 OPERATION IN FREEZING CONDITIONS ...................................................... O
3.4 SLIME/ALGAE GROWTH ..................................................................................O
3.4.1 PH ANALYSIS .............................................................................................P
3.4.2 NUTRIENT NITRATE AND PHOSPHATE ANALYSIS .......................................P
3.4.3 VISUAL INSPECTIONS ..................................................................................P
3.5 PERFORMANCE BENEFITS AND MAINTENANCE REQUIREMENTS................Q
3.5.1 METHOD..................................................................................................... Q
3.5.2 MAINTENANCE LOG ................................................................................... R
3.5.3 STREAM AND TURBINE FLOW .................................................................... R
3.5.4 PERFORMANCE OF TURBINE/EFFICIENCY .................................................. R
3.5.5 VISUAL INSPECTION OF TURBINE AND RUNNER ......................................... R
3.6 MECHANICAL INTEGRITY .............................................................................. R

4 COST BENEFIT ANALYSIS............................................................................... S

4.1 MAINTENANCE COSTS FOR A COANDA SCREEN AND CONVENTIONAL


SCREEN .........................................................................................................................S
4.2 TURBINE AVAILABILITY AND EFFICIENCY COSTS ....................................... T
4.3 CAPITAL COSTS .............................................................................................. T

5 SUMMARY OF RESULTS ................................................................................. V

5.1 SCREEN CAPACITY. ........................................................................................ V


5.2 SILT EXCLUSION PERFORMANCE. ................................................................. V
5.3 SELF CLEANING OPERATION AND PERFORMANCE IN FREEZING
CONDITIONS................................................................................................................ V
5.4 SLIME AND ALGAE GROWTH ........................................................................W
5.5 OPERATION AND MAINTENANCE REQUIREMENTS......................................W
5.6 INTEGRITY AND RESISTANCE TO DAMAGE ..................................................W
5.7 COST BENEFIT ANALYSIS ..............................................................................W

6 CONCLUSIONS ................................................................................................... X

7 RECOMMENDATIONS ..................................................................................... Y

APPENDIX A TYPICAL INSTALLATION DRAWING OF A COANDA


SCREEN ........................................................................................................................

APPENDIX B MINERALOGICAL REPORT OF CATCHMENT SILT


SAMPLE

APPENDIX C PHOTOS OF INSTALLATION .....................................................

B
APPENDIX D SITE DAILY RECORDS .................................................................

NB APPENDICES A-D AVAILABLE IN HARD COPY ONLY

C
2 TECHNICAL BACKGROUND

2.1 Description of Aquashear Screen


The Coanda effect static intake screen was developed in the USA to provide a
reliable maintenance free water intake screen for small hydro plant. A
drawing of a typical intake screen arrangement is attached (see Appendix A).

The screen is manufactured from an inclined wedge wire screen panel,


installed on the sloping downstream face of an over flow weir. At the top of
the screen is a curved acceleration plate. This can be of varying geometry, but
is often a simple circular arc. The objective is to smoothly accelerate the flow
over the crest and deliver the flow tangentially to the upstream edge of the
screen.

The flow to be abstracted passes through the screen into a conveyance channel
or chamber beneath the screen. Any excess flow, together with fish, larvae
and debris, is carried off the toe of the screen.

The screens are substantially self cleaning for most types of debris, due to the
high sweeping velocity down the screen face (usually greater than 2m/s). The
total head loss across the screen is approximately 1.3m, however, reduced
height screens that have a lower head loss can be supplied.

The screen has a very high flow capacity for a small aperture size, which
makes it possible to excludes leaves, fine debris, larvae and fish eggs from the
turbine system. The screen manufacturers literature states that all silt particles
greater that 1mm and 90% of silt particles between 0.5mm and 1.0mm2 will be
excluded.

The high capacity of the screen is achieved by a combination of the Coanda


effect - the tendency of a fluid to cling to a solid surface and the shear
produced by the sharp edges of the wedge wire. This is one of the key
features of the screen each wire is tilted, typically at an angle of five
degrees. In combination with the Coanda effect causing the fluid to cling to
the wire surface, the shearing action and hence capacity, is enhanced. The
screen forms the downstream side of a weir, and water passes down through it,
to the turbine penstock, or power canal, while silt and debris are washed off
the screen and back into the watercourse.

The Coanda screen has applications in many medium and high head hydro
schemes. The standard screen has a nominal flow capacity of 140 litres/sec per
linear metre. However, this is a conservative figure to allow for non ideal
flow entry to the screen and for some long term wear and reduction in

2
Suppliers data (Aquashear).

D
capacity. The maximum sized screen is generally limited by the physical
constraints of the site and economic considerations. Screens with a capacity of
up to 6 m3/s have been installed in the USA (for a hydro power site with a
head of 200m a flow of 1 m3/s equates to around 1.5 MW).

Currently the screen material is manufactured in the USA with the framing
and assembly undertaken in mid Wales. It is expected that as orders increase, a
larger proportion of the manufacturing will take place in the UK.

Dulas are introducing the screens to Europe and South Asia. The results of this
project will provide useful information for the application of the screen in
these areas. It will provide data on the capacity and effectiveness of the
screen, its affect on turbine performance and operation and maintenance costs.

2.2 Benefits of the Aquashear Coanda Screen

2.2.1 Silt Removal


The usual method of silt removal is the use of settling tanks. These can be
expensive to install and require frequent clearing out to maintain their
effectiveness. A settling tank can become filled very rapidly in a flood
situation and will lose its ability to remove silt particles from the turbine water
supply. Silt particles may cause turbine wear, which will cause a loss in
performance and result in increased maintenance costs, with subsequent loss
of output and revenue. A Coanda screen eliminates the need for a settling
chamber in most cases and significantly reduces maintenance costs, while
offering a much more reliable and robust desilting system.

2.2.2 Organic Material Removal


The removal of organic matter can enable a higher efficiency turbine to be
used, as well as removing the possibility of jet blockage. When a Pelton or
Turgo type turbine is used, a spider support is often used on the spear valve,
which can increase turbine performance, by acting as a flow straightener.
However, a spider support will be clogged by organic matter if it is not
removed from the water flow. The Coanda screen eliminates virtually all
organic matter with the potential to cause blockage in the turbine system.

2.2.3 Effect on Flora and Fauna


Environmentally the screen is very benign. There is no possibility of fish,
however small, being entrained, as with a conventional trashrack. Fish
generally pass unharmed over the screen, although some abrasion can occur if
there is insufficient flow to completely wet the whole screen. The screens
were extensively tested in the United States and approved by the Oregon
Department of Fish and Wildlife.

Any organic material in the stream will stay in the stream, being washed off
the face of the screen. This can be a financial as well as environmental benefit
as in some European countries any debris removed from a watercourse cannot

E
be returned to the watercourse and must be disposed of elsewhere, probably in
costly landfill. This practice is currently advised by the Environment Agency
and may in future become mandatory.

2.2.4 Potential Markets


There are many potential markets for the screen around the world. In
particular, in Indonesia and the Himalayas, where there are large silt problems
and many higher head sites, and in high altitude sites in Europe (the Alps and
Pyrenees) where freezing conditions and hard quartzite silt particles are
common.

2.3 Project Location


The Lodore Falls scheme is grid connected with a rated capacity of 170 kW.
The water flow taken by the Turgo impulse turbine is 235 l/s at full power and
there is an agreed minimum compensation flow of 140 l/s to remain in the
river.

The intake is located in a very inaccessible location with no electrical power


available. Any conventional intake, cleaning device and/or settling tank would
require operator maintenance, which is expensive and can be difficult in this
type of location. This is typical for many small hydro plants.

The scheme has been developed by Hydro Energy Developments Ltd, who
also operate the scheme. Prior to the installation of the Coanda screen, there
was an existing intake that suffered from blockage and lead to loss of power
and revenue, and also increased maintenance costs. Silt particles were not
excluded with the existing intake and would have been causing wear to the
turbine jet and runner.

Two new sections of Coanda screen were installed at the existing intake. The
screen is 2.13m wide in total and is nominally rated at 300 l/s capacity. The
two sections are 1.524m wide 0.608m wide. These were fitted into the
existing weir and header tank structure. The use of two sections enabled the
removal and examination of the smaller section after one year of operation.

F
3 METHODOLOGY OF ASSESSMENT AND RESULTS

The method of assessment of the key parameters relating to the performance


and cost benefit of the Coanda screen as a water intake, for hydro-electric
schemes, water supply or other purposes, is described below.

Screen capacity
Screen wear
Silt removal
Self Cleaning Operation
Operation in freezing conditions
Effect of slime/algae growth
Performance Benefits
Maintenance Requirements
Mechanical Integrity

3.1 Screen Capacity and Wear


The objective is to measure the screen capacity when newly installed and after
12 months use. The capacity of the screen is dependent upon the shape and
orientation of the wedge wires used for its construction. The sharp edges of
the wedge wires will become eroded over time and it is expected the flow
capacity of the screen will gradually reduce.

3.1.1 Method

Capacity
The capacity of the screen was measured shortly after commissioning and then
again after 15 months operation. The same method was used on both
occasions.

Water flow through the screen was measured using the ultrasonic flow meter
installed in the powerhouse, at the site.

The screen capacity is greater than that required by the turbine/pipeline


system. The large section of screen was covered in a plastic sheet, to reduce
the flow capacity below the turbine capacity, and the turbine flow then
gradually increased until all the water passing over the top of the small screen
was drawn into the turbine. The amount of water available to the screen was
controlled by adjusting temporary stop logs on the top of the weir wall. See
photos in Appendix C.

The maximum flow was determined by monitoring the pressure at the turbine.
As soon as the pressure started to fall, this indicated that the turbine was
taking more flow than the screen could supply, and that the pipe was
emptying. The turbine spear valve was then closed slightly until a steady
pressure was achieved. This was therefore the approximate screen capacity.

G
3.1.2 Results after commissioning
The measured width of screen left available for passing flow was 0.58m. The
maximum flow that could be taken by the turbine was recorded at somewhere
between 110 and 115 litres/second.

It is necessary to make some approximations to adjust this figure. These are


as follows;
1) It was not possible to completely seal the plastic sheet to the screen.
Hence there was a small flow of water that was able to pass beneath the
sheet and through the screen. This was small however, and was estimated
at not more than one litre/second.
2) There was also a small amount of water that was able to pass down the
side of the base plate at the foot of the screen, which would add to the
turbine flow. This was estimated at about two litres/second.
3) The stoplog arrangements at the top of the weir disturbed the flow entry to
the acceleration plate and screen, with the result that there was a small area
of flow separation, where the water did not contact the screen. It was
estimated that this resulted in a loss of screen area available of about 3%,
or three litres/second.

Hence the combined effect of these approximations;

Measured flow capacity = 112.5 l/s,


Losses due to un-utilised screen area = three litres/second,
Extra flow due to leakage around screen = three litres/second,

gives an estimated capacity of 112.5 litres/second for 0.58 m width.

This equates to a flow capacity of 194 litres/second per linear meter width.

3.1.3 Results after 15 months operation

The measured width of screen left available for passing flow was 0.49m.
This is a slightly different width to that used in the initial tests due to the
practicalities of arranging for exactly the same area of screen to be covered by
the plastic sheet. The maximum flow that could be taken by the turbine was
recorded at somewhere between 85 and 90 litres/second.

As before, there was some leakage beneath the plastic sheet and around the
plate sides, which was estimated in this instance to be a total of around 3 l/s.

There was a similar disturbed flow entry to the acceleration plate and screen,
with the result that there was a small area of flow separation (again about 3%),
where the water did not contact the screen. Additionally, in this test, there
was not quite enough water to ensure that the uncovered section of screen was
100% utilised. By using the stop logs to block off part of the screen, the water
level behind the weir was raised to the extent where flow started to spill over
the whole weir length, and was therefore not available to direct over the
uncovered section of screen.

H
It was estimated that the above two effects resulted in a loss of screen area
available of about eight percent, or seven litres/second.

Hence the combined effect of these approximations;

Measured flow capacity = 87.5 l/s,


Losses due to un-utilised screen area = seven litres/second,
Extra flow due to leakage around screen = three litres/second,

gives an estimated capacity of 91.5 litres/second.

This equates to a flow capacity of 187 litres/second per linear meter width
compared to 194 litres per second when measured after installation. This is
approximately a four percent reduction in capacity.

Errors

The measurements above are all subject to inaccuracy. Some estimate for the
degree of inaccuracy are indicated below:

1) The measurement of the point at which the screen is at full capacity is


probably subject to an error of plus or minus 2.5 litres/second or about
2.5% of total. This is mainly due to determining exactly when the pipe
begins to empty, and also that the turbine flow meter reading tends to
oscillate by plus or minus 1 litre/second.
2) The estimate of total leakage flows is probably plus or minus one
litre/second or about one percent of total.
3) The estimate of screen unutilised is probably plus or minus two percent.

The expected uncertainty is thus calculated by the square root of the sum of
the squares of the individual errors, ie.

Total Error = (2.52 + 12 +22) = 3%.

Hence the measured reduction in capacity is four percent, plus or minus three
percent.

3.1.4 Wear
The screen was examined for wear. A magnifying glass was used to examine
the edges of the wedge wires for comparison with the wires of a new screen.
It was not possible to see any general deterioration of the wedge wire profile,
although there were occasional scratches and slight deformations, probably
where larger stones had impacted.

What was noticeable however, was the thin coating of algae that covered parts
of the screen. Although not present upon the leading edge of each wire,
probably due to the shearing action of the water, it was present on the upper

I
surface of the wedge wire. This may have some effect on the capacity of the
screen, since the surface will be slightly roughened and this may reduce the
Coanda effect.

As a second method, a set of feeler gauges was used to determine the gap
between the wires of the installed screen and the gap for a section of new
screen for comparison. The measured gap varied between 1.0 and 1.1mm.
There is a slight error in measurement, since the wedge wires are able to
deform slightly. However, after measurement of the wedge wire gap in 20
different areas of the screens, there was no discernible difference between the
new and used screens.

3.1.5 Mineralogical Analysis of Silt


Samples of silt, which had collected above the weir, were taken as typical of
material that goes over the screen. There is a bias towards the collection of
larger particles as smaller particles tend not to be deposited.

The type of particles collected were analysed to determine the type of rock, ie
quartzite, sedimentary etc and to assess the abrasive properties.

The complete mineralogical report is contained in Appendix B.

The report states that the dominant material is made up of clasts of igneous
rock, with only a small percentage of sedimentary rock present. Quartz was
present but only as a minor phase. There is a high percentage of angular
shaped debris, indicative of the high energy environment. In abrasive terms,
the dominant rock had a hardness of around H6 on the Mohs Scale, H6 being
equivalent to Feldspar but softer than quartz. H5 and above will scratch
copper and H6 will scratch window glass. The screen material is stainless
steel, which has a hardness of between H6 and H7. We can conclude
therefore, that will be material in the river that can cause the screen to wear,
but as it is of a similar hardness, or perhaps slightly softer than the screen
material, we would expect the wear to occur over a long period of time. The
harder quartzite particles that could cause wear problems are only present in
small quantities. Hence the screen should be expected to have a long design
life under these site conditions.

3.1.6 Discussion
The combined results above show a slight deterioration in the screen capacity
after 15 months operation. However, the inherent errors involved in accurately
measuring such a small deterioration on a real site, make it difficult to draw a
definite conclusion. It is probable, given the results of the mineralogical
analysis above, that if there is a real reduction in capacity, it is probably more
due to algae build up, than wear of the wedge wires.

It shows that the practice of supplying schemes with screens that are 50%
higher in capacity is prudent. This will ensure that site specific effects such as
non ideal, (non laminar flow) inlet conditions, algae growth and screen wear,
will not affect the performance of the screen in the short or medium term.

J
What would be more useful, would be to revisit the scheme in three years time
and measure the screen capacity. This would indicate a more long term trend
in capacity change. It may also prove that the schemes are being supplied
with too much spare capacity, raising the potential for supplying smaller
screens at reduced cost.

3.2 Silt removal


The Coanda screen is designed to remove debris and silt particles from a river
flow and provide a supply of cleaned water to a hydro turbine.

The ability to remove silt particles is important in the long term operation of a
hydro intake system. Silt particles will cause wear to a turbine runner that will
reduce operational efficiency and increase maintenance costs.

3.2.1 Method
Graded sand was used to simulate silt particles in the water flow. Measured
quantities of sand of a known size distribution were poured over the top of the
screen with the turbine running. Any particles that passed through the screen
were collected by a fine muslin sheet temporarily installed beneath the screen.

Material collected in the muslin sheet was then dried, sieved and weighed to
give the particle size distribution not excluded by the Coanda screen.

The key performance parameters of the Coanda screen, as quoted by the


screen manufacturers, is that all particles above 1mm minimum diameter will
be excluded by the screen, and 90% of particles excluded down to 0.5mm
minimum diameter.

It was therefore apparent that the most interesting particle size range for
analysis is for particles greater than approximately 0.5mm minimum diameter.
A sample of sand was separated using three different square wire meshes to
give particle sizes between 0.41mm and 1.17mm. The masses of sand for
each particle size range, used in the test, are shown in the tables below.

3.2.2 Silt Exclusion Results

Particle Size Mass of Sample Mass of Sample % Exclusion


Range (S) Size Poured onto Size that Passed
(mm) Screen (g) through Screen (g)
1.17>S>0.70 200 91 55
0.41>S>0.70 100 82 18

Table 3.2.1 Low Flow Silt Exclusion, Test 1

K
Particle Size Mass of Sample Mass of Sample % Exclusion
Range (S) Size Poured onto Size that Passed
(mm) Screen (g) through Screen (g)
1.17>S>0.70 165 45 73
0.41>S>0.70 120 63 49

Table 3.2.2 Low Flow Silt Exclusion, Test 2

3.2.3 Analysis of Silt Exclusion Performance for Low Flow Tests

The results show that for the above two tests, a higher proportion of silt
particles between 0.41 and 1.17mm diameter passed through the screen than
predicted.

The difference between the two tests is probably explained by the rate of flow
used for each test. On the second test, the proportion of screen used was made
smaller by the use of wooden baffles, hence there was a higher flow for the
relative area used in the test. As a result, the second tests showed a higher rate
of silt exclusion.

The rate of flow is quite critical, since in reality, there will only be high silt
load when there is high flow. At high flows, there will be large excess flows
over the screen, hence the silt tends to be washed over the screen rather than
being allowed to settle on unwetted areas and then dropping through, as
happens under low flow conditions. At low flows, there is much less particle
transport, particularly of larger, heavier silt particles. Hence, a more
appropriate test would be under high flow conditions, as detailed below.

3.2.4 Higher Flow Tests

Although the flow available was not excessively high, it was found that by
making a temporary dam across the weir, the extra volume of water stored,
when released quickly, allowed a sufficiently higher flow across the screen.
Two tests were carried out with the following results;

L
Particle Size Mass of Sample Mass of Sample % Exclusion
Range (S) (mm) Size Poured onto Size that Passed
Screen (g) through Screen (g)
1.17>S>0.70 150 3 98
0.41>S>0.70 100 9 91

Table 3.2.3 High Flow Silt Exclusion, Test 3

Particle Size Mass of Sample Mass of Sample % Exclusion


Range (S) (mm) Size Poured onto Size that Passed
Screen (g) through Screen (g)
1.17>S>0.70 185 3 98.5
0.41>S>0.70 145 17 88

Table 3.2.4 High Flow Silt Exclusion, Test 4

These higher flow tests confirmed the expectations, with an average silt
exclusion rate of 94%. The excess flow and higher flow velocities that leave
the bottom of the screen does not allow particles to settle on the screen
hence they are not able to make the necessary change in direction to pass
between the screen wedge wires.

3.3 Self Cleaning Operation and Operation in Freezing


Conditions
The screen is designed to operate without any external intervention to remove
debris from the screen.

Coanda screens have been installed in other countries, in temperatures down


to 320C, without becoming frozen and blocked. This project has allowed
further verification of the performance of the screen under freezing conditions.

3.3.1 Method
The functional operation of the screen was monitored over a full year of
operation. This has included a winter season where cold temperatures were
encountered.

Monitoring of air and water temperature was carried out, using stand alone
single channel temperature recording loggers, designed for complete

M
immersion with internal temperature sensors. The logger data was downloaded
every month and an analysis of the data produced by Dulas.

The stream and turbine flows were logged for the duration of the project. If
the stream flow was high but the turbine was taking less flow, this could be
evidence of blockage of the screen (provided no operational or maintenance
procedures were responsible).

Photographs were taken of the screen at regular intervals to record any build
up of debris or signs of blockage. These are contained in Appendix C.

Appendix D contains the daily site records for Turbine flow, River Flow, Air
Temperature, Water Temperature, Power Output and Total kWh generated.

3.3.2 Operator Reports


In the 15 month period of operation, there has been no evidence of any
blockage of the screen due to leaves or other debris. The operator at the site
has been very pleased with the new screen, since the previous screen required
frequent cleaning, sometimes as much as every few hours during the autumn,
when there were lots of leaves present.

However, it was noted that over the last six months the turbine output power
had reduced by about 10kW, or approximately six percent. The turbine spear
valve was opened and examined and was found to be partly obstructed by
long, thin strands of weed, which had become hung up on the spear support
vanes. The spear was cleaned out and the turbine was returned to service.
The power then increased by about 7kW. The remaining 3kW of lost power is
due to fouling of the pipe which had become coated with a thin film of slime.
This has the effect of increasing the surface roughness of the pipe, slightly
decreasing the diameter, resulting in a higher friction loss for the pipe.

Hence the loss in efficiency due to weed passing through the screen, and
fouling the spear valve, is about four percent over 15 months.

Examination of the screen revealed a few strands of weed caught between the
wedge wires. It seems probable that this type of material may be able to find a
way through the screen. However, the concrete surfaces surrounding the
screen had become slightly eroded over time, hence there are some small gaps
that have formed around the edges of the screen. It is therefore probable that
some of the material has entered the system around the side of the screen, as
well as passing through the wedge wires.

The conclusion we can make from these observations, is that there may be
some types of organic material, made up of long thin strands, that can pass
through the screen and have the potential to cause some obstruction.
However, in this instance, this was a very minor problem and required only
about an hour to open the spear valve and clear the weed away. This problem

N
could be further reduced by ensuring that the screen remains well sealed to the
surrounding support structure.

The operator plans to monitor the power output closely to track any build up
of weed, and will plan in a 6 month inspection of the spear valve. This can be
planned for periods of low flow, when the turbine is shut down, hence this
would not involve any loss of production.

It should be emphasised, that compared to conventional alternative screens,


with larger apertures and less self cleaning characteristics, this is an almost
negligible problem. For conventional screens, it could be necessary to clear
the screen or turbine spear valve much more frequently, possibly every day at
certain times of the year.

3.3.3 Operation in Freezing Conditions


There was no air temperature data for the period from 12th April 1999 until the
end of August 1999. However, the water temperature during this period never
fell below 5 degrees Celsius, and it is unlikely that any frosts occurred during
this spring and summer period.

Both air and water temperature data were available from the period from 1st
September 1999 until 25th April 2000, which covered the autumn and winter
periods. Due to the relatively mild winter, there were no severe frosts;
however there was a continuous period of four days, from the 18th to 21st
December 1999, when the temperature remained between 0 degrees and -2
degrees Celcius. The screen continued to function normally throughout this
period and there was no evidence of any ice formation on the screen.

There were also other brief periods when the air temperature fell to freezing
point; no ice formation or reduction in screen capacity occurred.

Data obtained recently from other installations has shown resistance to much
more severe freezing conditions. Screens installed in other countries, notably
Switzerland3 and Utah, USA4, where temperatures have fallen as low as 32
degrees Celcius for prolonged periods, have not experienced any blockages
due to ice formation on the screen.

3.4 Slime/algae growth


One significant factor affecting the performance of the screen would be the
build up of large deposits of algae or slimes. The spacing of the screen
elements is only 1mm, hence the potential exists for these to become blocked.

Factors which may affect this are the pH (acidity) of the water and the
concentrations of the nutrients nitrate and phosphate in the river.

3
Graubunden, Switzerland. Capacity 140 l/s.
4
Beaver City, Utah. Capacity 740 l/s.

O
The sample of water collected from the river was analysed for pH, nitrate and
phosphate by the Institute of Freshwater Ecology at Windermere, Cumbria,
UK.

3.4.1 PH Analysis
Growth of algae can be promoted in pH neutral waters, whereas growth can
become inhibited in acidic waters (of pH 4-5 or less). The pH of the sample
collected at Lodore falls was 7.33. Hence, the pH of the water is conducive to
the growth of algae.

3.4.2 Nutrient Nitrate and Phosphate Analysis


The results are presented below, along with typical results for other waters in
the region, to allow comparison.

Parameter Lodore Falls Ullswater Coniston, Windermere,


(mg/litre) (Mildly Eutrophic) Haweswater
(mg/litre) (mg/litre)
Nitrate 135 1300 80-130
Soluble Reactive 2.5 30 10
Phosphate

Soluble Reactive Phosphate is that which is considered available for biological


assimilation, that is, that which would be available for algae or slime growth.

As a comparison, data was collected for Ullswater, which is considered mildly


eutrophic, and thus has higher than normal nutrient levels of nitrate and
phosphate. Windermere, Coniston and Haweswater all had similar levels of
phosphate but only about 10% of the nitrate concentration found in Ullswater.

It appears that the river at Lodore falls has average to low levels of nutrients,
compared to other waters in the region.

Given the results above, it may be expected that there would be some potential
for growth of algae. This is supported by the fact that the silt sample collected
was found to have a heavy coating of algae. This has probably built up over a
long time. However, excessive rate of algal growth due to nutrient
enhancement would not be expected.

3.4.3 Visual Inspections


From commissioning in April 1999 until around March 2000, there was a
limited build up of algae on the acceleration plate and upper part of the screen
material. There was no noticeable effect on performance. Since March
however, the rate of growth increased over the next few months. The operator
is of the opinion that this growth rate seems to be increased when there is
increased sunlight and lower flows. It appears that the build up of algae can
occur under these circumstances, but when the flow increases again, some
scouring takes place to limit further growth.

P
As a further example, at another site where a Coanda screen had been
installed, a small tree branch had become wedged between the wing walls of
the installation. It was noticed that there was a build up of algae in an area
underneath the branch. Although this area was subject to spray and splashing,
it was being protected by the branch from the scouring action of the water,
thus allowing algae to grow. The other unprotected areas of the screen
remained free of algae.

At the beginning of June, it was noted that the turbine flow had decreased
slightly, so the screen was inspected. It appeared that the top section of the
screen was partially blocked. The screen was cleaned using a stiff deck brush.
The algae was found to be easily removed and it was immediately obvious that
the screen capacity had increased, since the area of screen covered with water
decreased substantially.

Although not identified as a significant problem at the Lodore Falls site, it


would be prudent to recommend occasional monitoring of the amount of algae
on the screen. Other sites may lend themselves to higher deposits of algae,
depending on temperature, sunlight, orientation of the screen and water
quality. It may be necessary in some situations, to brush the screen every six
months or so, to maintain optimum performance.

3.5 Performance Benefits and Maintenance Requirements


Monitoring of power output and stream water level will indicate if the turbine
is taking all available flow or if flow is being restricted by a blocked intake.

3.5.1 Method
A log has been maintained of planned and unplanned maintenance on the
intake, turbine jet and runner. This will record details of the item of work and
the hours taken.

The logging of stream flow and turbine flow provides data as to whether the
turbine was taking its maximum allowable quantity of water and indicates if
there were any restrictions to the turbine water flow caused by a blocked
intake screen.

The performance of the turbine/generator system in terms of power output


against head and flow has been recorded at the beginning and end of the
monitoring period.

The condition of the turbine jet and runner has also been inspected at the end
of the monitoring period. This has been used to judge wear on turbine
components, which may be caused by silt.

Q
3.5.2 Maintenance Log
Any maintenance activities have been recorded on the daily log sheet,
contained in Appendix D. The log shows that there were no shut downs or
unavailability of the turbine due to screen or turbine blockage. The only
maintenance activities carried out are greasing of the turbine bearings and
exercising the valves. These activities are carried out when the turbine is shut
down due to insufficient flow, where possible.

3.5.3 Stream and Turbine Flow


The daily recordings given in Appendix D, show readings for turbine flow,
and river level. There is no evidence from these readings to suggest that the
screen was blocked at any time. A more reliable indicator is that there is an
operator on site every day and it would have been noted if there had been any
blockage of the screen or restriction of the flow to the turbine.

3.5.4 Performance of Turbine/Efficiency


Apart from the slight deterioration in performance due to weed on the spear
valve and sliming of the pipeline, as discussed in section 3.3.1, there has been
no deterioration in performance noted, since commissioning.

3.5.5 Visual inspection of turbine and runner


A visual inspection of the turbine runner has not revealed any signs of
abrasion or wear.

3.6 Mechanical Integrity


The screen and mounting should be able to withstand the forces imposed
during high flood conditions. The integrity and condition of the screen
mounting has been examined regularly, for evidence of damage after serious
flood conditions.

Since installation, the screen has been exposed to some very high flood flows,
with substantially sized debris (large rocks and at least two tree trunks) being
washed over it. There is no evidence of any damage to the screen. The main
reason for the high resistance to damage lies in the orientation of the screen.
The screen is on the down stream side of the weir and so any debris tends to
roll down the screen. There is no direct impact of rocks, branches or tree
trunks, hence less potential for damage.

R
4 COST BENEFIT ANALYSIS

As a result of the above analysis, a financial assessment has been made of:
the maintenance costs incurred with the Coanda screen and those
expected with a conventional screen installed. Two assessments
are made for the conventional screen. The first is for the screen
only. The second is for a screen and a desilting or settlement tank.
This allows a fairer comparison between the conventional solution
and the Coanda screen option, as it includes desilting performance.
the turbine performance in terms of availability and efficiency
the costs of avoided shutdowns due to non required maintenance (if
any).
the cost of supply and installation of a Coanda screen as compared
to a typical bar or mesh intake.

This results in a cost / benefit analysis for the installation and use of the
Coanda screen at this location.

4.1 Maintenance Costs for a Coanda Screen and Conventional


Screen
The maintenance requirements for the Coanda screen since commissioning
have been assessed in terms of costs. The only maintenance activities
required over the 18 month period was the cleaning of the spear valve and
removal of algae. It is anticipated that it will be a requirement to inspect, and
if necessary, clean the spear valve, about every 6 months. For this particular
site, where there is an operator on site, this will take approximately one hour
of labour, with no material costs. At an assumed labour cost of 15/hour, then
the maintenance direct cost is therefore 15 every six months or 30/year.

The screen was also cleaned once during the 15 month period to remove algae.
We can assume that at worst there may be up to two times a year where the
screen will need brushing. This would require a total of around four hours at
15/hour or 60/year.

Both the above procedures can be planned activities, hence there should be no
loss of production.

Hence the total estimated maintenance costs for the Coanda screen are
90/year.

The operator at the site, who has experience of the previous screen, reported
that it would be necessary to clean this screen as much as twice a day at
certain times of the year, when there was higher amounts of vegetation and
organic matter in the river. This may not be typical for other alternative types
of screen, however, cleaning on an average of twice a week would not be

S
unexpected. Hence, on this basis, typical costs would be of the order of two
hours per week at 15 per hour, ie 30 per week or 1560/year.

The desilting tank would also need regular maintenance, to periodically flush
out the accumulated silt. It is estimated that this would require about 2 hours
every month, ie 24 hours per year, resulting in a cost of 360 per year.

4.2 Turbine Availability and Efficiency Costs

The loss of 7kW at full power, due to the weed around the spear valve would
not have occurred under normal circumstances. The reason it was not
removed earlier was because the problem was thought to be due to sliming of
the pipeline, which is much more expensive maintenance task, and therefore
cannot be undertaken so regularly. Now that this problem has been
recognised, monitoring of the turbine output power will reveal weed build up
at an early stage, and this will be easily dealt with.

Hence we can conclude that there will be no costs associated with loss of
efficiency, due to screen problems. There have been no instances when the
scheme could not be operated due to screen blockages or maintenance, since
any limited cleaning required can be done at times of low flow, as part of a
planned maintenance schedule.

The operator on site estimates that compared to a manually cleaned


conventional screen, the installation of the new screen has saved around
100,000 units over the period of the project. Most of this saving is achieved in
the period from September to January where higher flows are normally
available but large amounts of leaves are capable of reducing a conventional
screen capacity by around 50%5. This loss of production occurs because with
high debris concentrations it is practically impossible to clean the screen
frequently enough to maintain generating capacity.

The selling price per unit is 0.04, hence the annual saving is 4000.

4.3 Capital Costs

The typical capital costs of a manually cleaned conventional screen compared


to a Coanda screen are very site specific. For example, where the Coanda
screen eliminates the need for a desilting basin or settling tank, it may be a
similar cost or even cheaper to install the Coanda screen. This is particularly
relevant in the developed countries, where labour rates and large civil works
are expensive. Even in developing countries this can also be true, as recently
illustrated by the installation of a 1200 l/s capacity Coanda screen in Sri
Lanka.

5
Based on data from other Hydro Energy Development Ltd. sites

T
At other sites, where silt abrasion and wear are not considered such a problem,
or civil works are very cheap, the Coanda installation costs will be higher.

Hence for this analysis, two capital costs have been compared for the
conventional screen solution. The first is for a conventional screen only. For
this particular site, with no desilting tank, the capital costs for a conventional
screen would be about 50% of that of the Coanda screen, based on screen
material costs and associated civil works costs. This is probably typical for
other UK sites.

The second cost is for a conventional screen plus settling tank, to allow a fair
comparison of like with like, in terms of desilting performance.

Hence, for this installation, the cost benefit analysis is as follows:

Conventional Conventional Coanda Screen


Screen Screen with
Settling Tank
Capital Cost 7400 14,000 14,800
Generated Units 700,000 700,0000 800,000
Generated 28,000 28,000 32,000
income/year @ 4.0
pence /kWh.
Maintenance Cost 1560 1900 90
NPV over 5 years1 94,000 86,000 108,000
NPV over 10 years1 169,000 161,000 199,000
1
Based on 2.5% annual inflation and 8.5% discount rate.

The results show that for this site, over a typical 10 year life span, the net
present value of choosing a Coanda screen is 30,000 higher than that for a
conventional screen and 38,000 higher than that for a conventional screen
with desilting tank. Even over a very short five year life span, there is a
saving of 14,000 and 22,000 respectively . This would be enough to build
the screen again in totality, including design and all civil works. In reality, in
the unlikely event that the screen needs replacement after only five years, the
replacement cost would only be 7200 (at 2.5% inflation), since only the
screen material would be replaced. (The surrounding support structure and
civil works would be designed for 25 years.)

Hence it would be possible to replace the screen every five years and still
show between 7000 and 15,000 saving each time.

U
5 SUMMARY OF RESULTS

The following summarises the results of each of the tests and analyses carried
out.

5.1 Screen Capacity.

The capacity of the screen was found to be approximately four percent less
than when installed. However, the errors present in accurately measuring the
flow were estimated at around 3%, hence it is difficult to draw a definite
conclusion. Microscopic examination of the screen wire edges showed no
visible signs of wear. The mineralogical analysis of typical silt found in the
river system, did not show any significant presence of highly abrasive
material. Hence it is thought unlikely that there would be a significant
deterioration of performance due to wear of the screen material.

There was however, evidence of the build up a limited amount of biological


growth, probably algae. This growth had coated part of the screen and it is
possible that this could reduce capacity slightly. The effects of this growth
can be mitigated by cleaning, and this is discussed more fully below.

5.2 Silt Exclusion Performance.

Several tests were carried out over a range of flows. At low flows, where
there was insufficient water to produce an excess leaving the bottom of the
screen, the silt exclusion performance was lower, averaging around 50%
exclusion of particles between 0.41 and 1.17mm. At higher flow tests, much
better performance was achieved, with average exclusion rates for the same
particle range of 94%. It was felt that this was a more indicative test, since
large volumes of small size silt particles would only be in suspension at high
flows. In such instances, there is sufficient excess flow to remove silt
particles from the screen, rather than allowing them to sit on the surface and
trickle through the wedge wires.

5.3 Self Cleaning Operation and Performance in Freezing


Conditions.

Neither the flow records for the turbine, or the operator reports showed any
evidence of blockage of the screen by debris. The operator reported that he
had never had to remove any debris or blockage from the screen.

It has been discovered however, that very thin strands of weed are able to pass
in small quantities through the wedge wires. This problem manifested itself in
a reduction in turbine maximum output power of about 7kW, or about four
percent. This occurred very gradually over a period of 15 months until the
operator decided to investigate the problem. Upon opening up the spear valve,

V
thin strands of weed were discovered. After cleaning these out, the turbine
output returned to normal. This was a short maintenance operation that only
required about an hour to carry out.

The winter of 1999-2000 was not particularly cold, so the screen was not
tested under severe freezing conditions. However, there was a few occasions
when the air temperature reached freezing point, and a period of several days
where it remained between zero degrees and minus two degrees celsius. No
problems of freezing of the screen or blockage by ice were noted.

5.4 Slime and Algae Growth

During the 15 month period, it was noticed that a thin film of algae had started
to spread over the screen. After about 12 months of operation, it was noted
that this had reduced the capacity of the screen. The screen was cleaned with
a stiff brush and the capacity of the screen returned to normal. Again this was
an isolated maintenance operation that only required about 1 hour to perform.

The water quality analysis showed that although the pH of the water was
conducive to algae growth, the concentrations of the nitrate and phosphate
were low, hence, higher than normal growth rates of algae would not be
expected.

5.5 Operation and Maintenance Requirements.

Apart from the two occasions described above, no maintenance was required
on the scheme due to problems with the screen.

5.6 Integrity and Resistance to Damage

The screen was inspected after several high floods where large tree trunks and
boulders had been washed over the weir. There was no evidence of any
damage to the screen. This is mainly due the orientation of the screen, where
objects tend to roll over it rather than impact with high force. This resistance
to damage has been observed at other installations, after more than 5 years in
service.

5.7 Cost Benefit Analysis

The NPV calculations carried out for the Coanda screen and a conventional
screen showed a significant advantage to using the Coanda screen. Over 5
years, there was a 14,000 benefit and over 10 years this amounted to
30,000. The majority of this benefit arises from the extra energy production
possible by using a self cleaning screen, that never blocks, compared to a
conventional screen can be 50% blocked for the whole of the autumn, when
high leaf loads are present.

W
6 CONCLUSIONS

From the results obtained from this investigation, it is clear that the
installation of a Coanda screen at this location has been highly successful,
resulting in payback on the installation costs within 1-2 years. The screen has
been effectively 100% reliable, since there has been no lost energy production
due to blockage by debris.

The screen has proven to be of consistent high capacity, robust, resistant to


damage from large boulders and tree trunks, unaffected by freezing
temperatures and has a high silt exclusion performance. The capacity tests
showed no significant deterioration in performance over a 15 month period.
The screens are also supplied approximately 50% oversized, hence even in
more aggressive, abrasive environments, the screens will be capable of
exceeding the performance requirements over a long design life.

Maintenance requirements have been negligible, with a total of about 3 hours


over 15 months being required for screen related maintenance activities. The
investigation has however, highlighted two minor areas where more careful
monitoring of performance may be required. The first is monitoring of algal
growth, which, if allowed to build up over long periods, can reduce the
capacity of the screen. In nutrient enriched waters, it may be necessary to
clean the screen once a month. The second area is monitoring of turbine
power output to check for build up of weed that has passed through the screen
and collected on the turbine spear valve. Both of these processes occur
relatively slowly, hence the cleaning can be a planned maintenance activity,
during a dry period. Hence, no loss of electricity production need occur.

Over the long term, the installation of an Aquashear Coanda screen is an


extremely good investment. For this particular site, the projected extra Net
Present Value from installation of the Coanda screen is 30,000 over a 10 year
period. For sites where a settling tank would be required in addition to the
conventional screen, the extra Net Present Value over 10 years was 38,000.
On some schemes, this extra availability and generated revenue, provided by
the Coanda screen, could be sufficient to turn an economically unattractive
project into a commercial viability.

X
7 RECOMMENDATIONS

The screen has been tested over a period of 15 months for a range of
performance parameters. The screen has performed well in all of these tests.
The conclusions for most of the tests can be easily and accurately extrapolated
to the long term. For instance, tests for performance under freezing
conditions, resistance to blockage by debris, impact damage resistance and silt
exclusion should not be time dependent. However, there are two performance
parameters that would be of interest to investigate over the longer term;

1) Reduction of capacity due to build up algae or biological growth. It would


be very informative to monitor the build up of algae after a longer period
of 3 or 5 years. During this time it would be expected that the screen
would continually accumulate algae, which would be removed at intervals
by the operator as necessary. The current method is to use a stiff brush to
remove the growth from the outside screen surface. An increase in
capacity back to normal is usually noted. However, this brushing does not
return the screen to bare metal condition. It is possible that over longer
periods, more resistant residues may accumulate which could require a
more intensive cleaning process.

2) Reduction in capacity due to abrasion and blunting of the wedge wire


edges. The river in which the screen was installed did not contain large
volumes of abrasive material. A more meaningful test to determine long
term performance and life, would be to test the capacity of the screen after
5 years. An alternative would be to carry out a perfomance monitoring
programme on a screen installed in a more silty, abrasive environment. A
screen has recently been installed in Sri Lanka, which experiences high silt
loading. As the site is also in a tropical climate the rate of growth of algae
is much faster than for sites in the UK. This could therefore be a useful
site to monitor these two parameters of abrasion and algae growth.

The Aquashear Coanda screen is also supplied with smaller wedge wire
spacings. It would therefore be interesting to test a screen with 0.5mm
aperture. This type of screen would have many applications for higher head
sites, where the removal of smaller silt particles is essential to minimise
abrasion. It would be possible to do this on a small scale, by installing a
small, prefabricated section of screen in an appropriate river and monitoring
algae growth, abrasion and capacity over a 3 or 5 year period. The screen
would not necessarily have to be part of a hydro scheme to test these
parameters.

Vous aimerez peut-être aussi